272 Profs. Dewar and Fleming on the Electrical 



with temperature below —100° C, and to determine the form 

 of the temperature-resistance curves as the temperature is 

 lowered down to the lowest points obtainable by the evapora- 

 tion of liquid oxygen boiling under reduced or normal pres- 

 sure. The two variables with which we are concerned being, 

 therefore, Specific Electrical Eesistance and Temperature, we 

 may point out in the first place how each of these has been 

 defined and measured in tabulating the results. 



§ 2. There are, as is well known, two ways in which we 

 may define Specific Electrical Resistance. The first, called 

 the Volume-Specific Eesistance, is defined as the resistance 

 of one cubic unit of the material taken between opposed faces 

 of the cube. In this method of defining it the specific elec- 

 trical resistance is given as the resistance in ohms, or electro- 

 magnetic units of one cubic centimetre of the material between 

 opposed faces. The second, called the Mass-Specific Resist- 

 ance, is the resistance of a wire of the material of a certain 

 length and weight. In this second method of defining it, it 

 is generally stated as the resistance in ohms of a wire of the 

 material one metre in length and weighing one gramme. 

 When it is desired to compare together or state the specific 

 resistance of materials in the form of wires, of which the dia- 

 meter is not very different from one millimetre or larger, and 

 length, say, of one metre or more, there is no doubt that 

 the most convenient way of stating the results is by means of 

 the mass-specific resistance ; and since the mean specific 

 gravity of wires of this size can always be determined with a 

 very fair degree of accuracy, the volume-specific resistance 

 can easily be deduced. If, however, as in the case of our 

 own experiments, the conditions necessitate the employment 

 of very fine wires of about '08 to *25 millim. in not very 

 great lengths, the probable errors on determining the weight 

 per metre, and especially the specific gravity, are very con- 

 siderable. We therefore preferred to face the difficulty of 

 determining the mean diameter of the wires employed and 

 state our results in terms of the volume-specific resistance. 

 One part of the work consisted, therefore, in the determina- 

 tion of the mean cross section of a series of very fine wires 

 prepared from the metals and alloys under test, taken by 

 means of an optical method to be presently described. From 

 this mean cross section and the length we are able to state, 

 when the resistance is known, the volume-specific resistance 

 of the material. Since, however, the measurements of length 

 and diameter were all taken at temperatures between about 

 15° C. and 20° C, it will only be possible to express the true 

 volume-specific resistance at any assigned temperature when 



